Three-Dimensional Structure of N-Terminal Domain of DnaB Helicase and Helicase-Primase Interactions in Helicobacter pylori

Replication initiation is a crucial step in genome duplication and homohexameric DnaB helicase plays a central role in the replication initiation process by unwinding the duplex DNA and interacting with several other proteins during the process of replication. N-terminal domain of DnaB is critical for helicase activity and for DnaG primase interactions. We present here the crystal structure of the N-terminal domain (NTD) of H. pylori DnaB (HpDnaB) helicase at 2.2 Å resolution and compare the structural differences among helicases and correlate with the functional differences. The structural details of NTD suggest that the linker region between NTD and C-terminal helicase domain plays a vital role in accurate assembly of NTD dimers. The sequence analysis of the linker regions from several helicases reveals that they should form four helix bundles. We also report the characterization of H. pylori DnaG primase and study the helicase-primase interactions, where HpDnaG primase stimulates DNA unwinding activity of HpDnaB suggesting presence of helicase-primase cohort at the replication fork. The protein-protein interaction study of C-terminal domain of primase and different deletion constructs of helicase suggests that linker is essential for proper conformation of NTD to interact strongly with HpDnaG. The surface charge distribution on the primase binding surface of NTDs of various helicases suggests that DnaB-DnaG interaction and stability of the complex is most probably charge dependent. Structure of the linker and helicase-primase interactions indicate that HpDnaB differs greatly from E.coli DnaB despite both belong to gram negative bacteria.


Introduction
Helicobacter pylori (H. pylori) is a gram-negative, microaerophilic spiral shaped bacterium that infects more than 50% of the human population globally and is responsible for causing chronic gastritis, peptic ulcer and gastric cancer [1]. Replication system in H. pylori has unique features compared to other well studied microorganisms like E. coli and Bacillus subtilis such as the absence of the recF gene, the presence of the dnaA gene ,600 kb away from the dnaN-gyrB genes and the absence of the dnaC gene [2].
Helicases are nucleic acid motor proteins that derive energy from NTP hydrolysis and unwind DNA in 59 to 39 direction [3]. In E. coli, DnaB helicase plays a central role; which displays proteinprotein interactions right from the beginning to the end of replication process. The partners of DnaB include but not limited to DnaC (loading partner), DnaA (initiator) [4], DnaG (primase) [5], SSBs (Single Strand Binding proteins) [6], t-subunit of DNA polymerase [7], and replication termination protein [8]. The loading partner DnaC recruits DnaB as DnaB 6-DnaC 6 complex at oriC site and plays an important role during replisomal complex formation [9,10]. Interestingly, bioinformatics studies shows the absence of a DnaC homolog in the H. pylori genome [2,11] and yeast two-hybrid studies shows that H. pylori DnaB (HpDnaB) does not interact with HpDnaA [12]. Our biochemical studies earlier had confirmed that HpDnaB is able to complement the DnaC function and loads itself to the oriC site to form the pre-replication complex [13].
Helicases are classified into three super families (SF1, SF2 and SF3) and one small family (F4) based on the presence of specific motifs [14]. The HpDnaB belongs to family 4 that consists of homohexameric replicative helicases. Each monomer consists of a N-terminal domain linked via a flexible linker to the C-terminal domain. The C-terminal domain is highly conserved as all the five conserved motifs including the ATPase motif and DNA binding motif are located within this region. The N-terminal region is highly variable among the helicases, similar to its binding partner, the C-terminus of DnaG primase.
Electron microscopic studies of the full length structure of DnaB helicase from E. coli (EcDnaB) [15], RSF1010 helicase [16], Bacillus subtilis bacteriophage SPP1 helicase (G40P) [17] and bacterio-phage T7 gene 4 DnaB [18] reveal a closed ring hexameric state with several different quaternary states like 6-fold rotational symmetry (C 6 ) or 3-fold rotational symmetry (C 3 ) or an intermediate between the two (C 3 C 6 ). Very recently the crystal structures of Bacillus subtilis bacteriophage SPP1 G40P helicase (G40P) [19], Bacillus stearothermophilus DnaB (BstDnaB) [20], Thermus aquaticus DnaB (TaqDnaB) [21] and the N-terminal domain structure of Mycobacterium tuberculosis DnaB (MtbDnaB) [22] have also been reported. The crystal structures have illustrated that the linker region between the globular N-terminal domain and the C-terminal domain consists of three a-helices, where first two helices forms a-hairpin structure. The structure of BstDnaB in complex with primase C-terminal domain, along with the structural and mutational studies on G40P helicase revealed that the N-terminus of helicases plays a crucial role in binding to primase and triggers primase mediated enhancement of helicase activity [19,20].
Our recent biochemical findings suggest that the N-terminus and linker region plays an important role in the multimerisation, quaternary state transition and activity of HpDnaB [23]. The DelN1HpDnaB deletion mutant (1-66 residues deleted), which lacks the NTD but consists of an intact a-helical hairpin and the DelN2HpDnaB deletion mutant(1-133 residues deleted), which lacks the first helix from a-helical hairpin and the NTD show 20% and 30% reduction in helicase activity respectively. The electron microscopic studies emphasized the essential role of N-terminal region for the quaternary state polymorphism in HpDnaB [23]. Similarly, the N-terminal domain appears to be critical in EcDnaB, BstDnaB and G40P helicase since an equivalent NTD deletion mutants consisting of the C-terminal helicase domain and a part of the linker region exhibit in vitro ATPase activity but fails to demonstrate any helicase activity [24][25][26]. However, the importance of NTD and linker for helicase activity does not explain the function they play during DNA unwinding process at the replication fork.
Replicative helicases bind DnaG primase to synchronize DNA unwinding with RNA primer synthesis during the process of replication [5]. DnaG consists of an N-terminal Zinc binding domain (ZBD), RNA polymerase domain (RPD) and a C-terminal Helicase binding domain (HBD). In different organisms, the DnaB-DnaG complex stability varies remarkably. In E. coli, the interaction is relatively weak [27] whereas in B. stearothermophilus, the DnaB-DnaG complex is highly stable as can be purified intact by gel filtration [24]. Based on these observations, we were interested in studying the DnaB-DnaG interactions in H. pylori.
Taking leads from the homology modeling together with in vitro and in vivo function of different deletion mutants of HpDnaB as shown earlier [23], we defined the N-terminal and C-terminal domains separated by the flexible linker region. We further cloned deletion constructs of N-terminal domain (NTD) consisting of 1 to 121 residues and N-terminal domain with longer linker consisting of 1 to 144 residues (NTDL). To understand the roles played by NTD in helicase function, the importance of the linker region and helicase primase interactions, we present here the structure of NTD of HpDnaB at 2.2 Å resolution and compare it with other NTD structures. Further, we were interested to see whether helicase-primase interaction is conserved in slow-growing pathogenic bacteria like H. pylori and whether it can account for the stimulation of key functions of these proteins at the replication fork. We report here the cloning, expression and purification of H. pylori primase (HpDnaG) and C-terminal domain (CTD) of HpDnaG primase (HpDnaG) for the very first time. We had investigated the role of HpDnaG as a possible binding partner of HpDnaB and its effect on helicase activity.

Cloning, expression and purification of NTD of HpDnaB Helicase
The gene corresponding to NTD of HpDnaB construct was amplified by Polymerase Chain Reaction (PCR) using wtHpDnaB construct as a template. The PCR amplification of 1-363 nucleotides was done using Taq DNA polymerase (MBI Fermentas) with the following forward and reverse oligonucleotide primers: FP: 59 CATGCCATGGATCATTTAAAGCATTTGC 39 and RP: 59 CCGCTCGAGGGCTTGCTCTCTAATGG 39 The amplified DNA was inserted into NcoI/XhoI digested expression plasmid vector pET 28a (Novagen, Madison, WI, USA) with C-terminal six residue His tag.
A little longer construct of NTD (NTDL) of HpDnaB Helicase was made to include helical hair pin region. The PCR amplification of 1-144 amino acids was done using Taq DNA polymerase (MBI Fermentas) with the forward and reverse oligonucleotide primers, FP: 59 CATGCCATGGATCATT-TAAAGCATTTGC 39 and RP: 59 CCGCTCGAGGCCATTCAATAACGCATAG 39 The amplified DNA was inserted into NcoI/XhoI digested expression plasmid vector pET 28a (Novagen, Madison, WI, USA) with C-terminal six residue His tag.
The recombinant plasmid containing the NTD of HpDnaB insert was transformed into BL21 (DE3) cells. Freshly transformed BL21 cells were grown in 2% LB medium containing 50 mM kanamycin at 37uC to an O.D 600 of 0.4. Then, the culture was induced with addition of 1 mM IPTG. Cell growth continued till O.D. 600 reached 0.9 and cells were centrifuged at 5000 rpm. The harvested cells were resuspended in lysis buffer containing 50 mM Tris.HCl pH 8.0, 0.3 M NaCl, 1 mM Imidazole, 100 mM Phenyl Methyl Sulphonyl Fluoride (PMSF) and lysed by addition of lysozyme (1 mg/ml) that was then incubated for ten minutes at 37uC. The clarified lysate was passed through a Ni-NTA column. The bound protein was eluted with 300 mM Imidazole (50 mM Tris, 150 mM NaCl, 0.3 M Imidazole, pH 7.5). The positive fractions were pooled, concentrated and purified further on a Hi Load superdex-75 16/60 column (GE) equilibrated with 20 mM Tris.Cl buffer pH 7.5 containing 100 mM NaCl, 1 mM EDTA, 5 mM bME ( Figure 1A). The fractions were loaded on 10% SDS-PAGE ( Figure 1B) and pure fractions were pooled and concentrated to 10 mg/ml in YM3K centricon tubes (Amicon). The concentrated protein was used for crystallization and biochemical assays. The full length HpDnaB and DelN2HpDnaB were expressed and purified as described before [23]. The expression and purification of NTDL was similar to NTD of HpDnaB as mentioned above.
The PCR product was subsequently digested with SalI restriction enzyme and cloned in pET30b (Novagen, Madison, WI, USA) expression vector at SalI site. E.coli strain BL21 (DE3) was transformed with pET30b-HpDnaG construct and grown at 37uC in LB medium containing 50 mM kanamycin until the O.D. 600 reached 0.6. Induction of His 6-HpDnaG protein (HpDnaG) was carried out using 1 mM IPTG at 22uC for 4 hrs. The bacterial cells were centrifuged to get the pellet, and the pellet was frozen in 280uC. Protein purification was carried out using Ni-NTA agarose beads (Qiagen, Hilden, Germany) according to manufacturer's instructions. The eluted protein was dialyzed against 20 mM MES, 100 mM NaCl, 10% glycerol, 1 mM EDTA, 100 mM PMSF and 10 mM bME. Second step of purification was carried out by ion-exchange chromatography using resource S ion-exchange column (GE Healthcare, Uppsala Sweden) in accordance with the manufacturer's instructions. The peak fractions were checked in 10% SDS-PAGE and dialyzed against 50 mM Tris.HCl pH 7.5, 100 mM NaCl, 10% glycerol, 100 mM PMSF, and 10 mM bME.
The CTD of DnaG Primase was amplified by using Taq DNA polymerase (MBI Fermentas) for PCR with genomic DNA of H. pylori (ORF Hp0012) as template. Following oligonucleotide primers were used as forward and reverse primers: FP: 59 CTAGCTAGCGAGCGAGTCTCTTTTCAGCCTTT 39 and RP: 59CGGCTCGAGTATGGCGACTAATTCTCCTTGT-TTT 39 The amplified PCR product was digested with Nhe1/Xho1 and inserted into Nhe1/Xho1 digested expression plasmid vector pET21c (Novagen, Madison, WI, USA). The CTD of HpDnaG with C-terminal six residues His tag was expressed and purified similar to the NTD of HpDnaB.

Crystallisation and data collection of NTD of HpDnaB
The purified NTD and NTDL of HpDnaB (10 mg/ml) was subjected to various crystallization screens (Hampton research) both at 16uC and 4uC. Hanging drops were prepared in 24-well plates by mixing 3 ml protein solutions (10 mg/m1) with 3 ml reservoir solution and were equilibrated against 500 ml reservoir solution. Crystalline precipitates were observed for NTD of HpDnaB with 25% PEG 3350, 0.2 M Ammonium sulfate and 100 mM Bis-Tris pH 5.5 as precipitant during the initial screens. After optimization of the physico-chemical parameters affecting crystallization, the best conditions were obtained at 4uC using the hanging-drop vapor diffusion method ( Figure 1C). So far, the crystallization trials on the NTDL were not successful. To provide cryo-protection, crystals of NTD of HpDnaB were treated with mother liquor containing 30% PEG400. The X-ray diffraction experiments were performed at 100K from crystals mounted in cryoloops at cryogenic conditions and flash frozen in liquid nitrogen. The X-ray diffraction data were collected at BESSY synchrotron beam line at wavelength of 0.9184 Å . The data set were indexed and scaled with program HKL2000 [28] or XDS [29].

Structure determination and refinement
The structure of NTD of HpDnaB was solved by molecular replacement using Phaser [30] with the structure of N-terminal domain of Mycobacterium tuberculosis (PDB id: 2r5u; Biswas et al, 2007) as a search model. For refinement, initially Refmac5 [31] was used and further improvements were done by CNS [32]. Along with two molecules of NTD of HpDnaB in one asymmetric unit, a large extra electron density was observed near the dimeric interface. This extra density showed well connected main chain and MALDI results ( Figure S1) allowed us to build the side chains. The structure was improved with iterative model building using COOT graphics package [33], the final structure had very good electron density ( Figure 2B). Water molecules were picked up both manually as well as with appropriate routines in CNS. The PyMOL program [34] was used to generate the figures. The R_factor and R_free values for this structure are little above the average values for the structures refined at given resolution [35], this could be because the crystals showed very high solvent content (77%) and some non traceable extra electron density, which could be due to low occupant degraded peptides. Helicase assay Helicase assay substrate was prepared as described earlier by Sharma et al [36]. Helicase assay was carried out in 20 ml reaction mixture containing 20 mM Tris.HCl pH 8.0, 8.0 mM DTT, 2.5 mM MgCl 2 , 2.0 mM ATP, 80 mg/ml BSA, 10 mM KCl, 4% sucrose, 10 fmol of helicase substrate and the indicated amount of HpDnaB and HpDnaG proteins ( Figure 3B). The reaction mixture was incubated at 37uC for 30 minutes. The reaction was stopped by addition of 5 ml 5X stop buffer (1.25% SDS, 75 mM EDTA, 25% glycerol) and loaded on 10% native PAGE to resolve the products. The gel was dried and exposed to X-ray film (Kodak) and developed subsequently.

Surface Plasmon Resonance (SPR) experiments
All binding experiments were performed on a Biacore 2000 apparatus (Biacore, GE Healthcare) at 25uC in running buffer (10 mM Hepes pH 7.4, 150 mM NaCl, 3 mM EDTA, and 0.005% P20 surfactant). A total of 750 resonance units (RU) of CTD of HpDnaG primase (ligand) were immobilized on a research grade CM4 chip (Biacore TM ) in 10 mM sodium acetate, pH 5.0 according to the manufacturer's amine coupling kit. The unreacted residues on the surface were blocked by two washes of 1 M ethanolamine, pH 8.5. One flow cell left blank was used as reference. Helicase samples (10 mM) were prepared in running buffer and were injected at 20 ml/min across the sensor surface. Signal changes on the activated/blocked control panel were subtracted from the DnaG-DnaB binding interactions using in-line reference signal and the subtracted sensorgrams were analyzed. The surface was regenerated with regeneration buffer consisting of a pulse of 25% ethanol. This experiment was repeated for full length helicase HpDnaB, DelN2HpDnaB, NTDL and NTD of HpDnaB. Data analysis was performed with the BIAevaluation software (Pharmacia, version 2.1). The dissociation and association rate constants k diss and k ass respectively were determined by direct curve fitting following the equation , where R t is the relative response at time t and R 0 is the relative response at the starting time t 0 . The association rate constants were calculated from the measured k diss according to the equation Here, R eq is the steady-state response level and C is the molar concentration of the analyte. The dissociation constant (K D ) were calculated by dividing k diss by k ass .

Protein structure accession number
Atomic coordinates and structure factors have been deposited in RCS Protein Data Bank (PDB) with accession code 3GXV.

NTD of HpDnaB exists as a dimer in solution
The NTD (1-121 residues) of HpDnaB was expressed and purified as described in Materials and Methods. The elution profile of size exclusion column shows that the protein exists in a dimeric form ( Figure 1A). The protein was isolated after gel filtration to homogeneity as estimated by SDS PAGE (Figure 1B). Rhombohedral crystals (0.1560.1560.1 mm) grew overnight against a crystallization well solution containing 0.5 ml reservoir solution containing 5% PEG 400, 150 mM ammonium sulphate and 100 mM Bis-Tris pH 5.5 ( Figure 1C). Unfortunately, NTDL (1-144 residues) did not form crystals under our experimental conditions.

Data collection and structure determination
Initial attempts for crystallization yielded small, rhombohedral shaped crystals that appeared overnight at 4uC. Several crystals were tried for data collection and few crystals diffracted with pseudomerohedral twinning were processed in tetragonal space group but refinement did not yield satisfactory R-factors (data not shown). Following this, a wide variety of crystallization conditions explored. Fortunately, the crystals appeared in the conditions mentioned in Materials and Methods, which diffracted to 2.2 Å resolution in orthorhombic space group without twinning. The best structure solution was obtained with recently released structure of NTD of Mycobacterium tuberculosis helicase (NtMtbDnaB, PDB iD: 2r5u), with two molecules in one asymmetric unit. The data processing and refinement statistics are in Table 1 and a portion of 2F o -F c electron density showed in Figure 2B.

Overall Structure
The NTDs of helicases are relatively less conserved (about 25% identity) ( Figure S2) compared to the C-terminal ATPase domain (about 40% identity) and these two domains are connected by a much less conserved linker region. We determined the structure of NTD of HpDnaB which is highly helical with a compact head (Nglobe) and a long helix (H7) at the C-terminal end (Figure 2A). The asymmetric unit consists of two complete monomers and two short peptide fragments. Interestingly, the two complete monomers form a dimer which is stabilized by the two short peptides and it eventually appears as a heterotetramer ( Figure 2C). In a dimer, the two monomers are arranged in a head-to-head manner where the C-terminal helices are extended parallel to each other. The two short peptides form helices, which are arranged in an anti-parallel fashion with the extended C-terminal helices (H7) resulting to the formation of a four helix bundle. The head-to-head dimer organization as seen in NTD of HpDnaB is different from other homologous structures known so far. The structure leads to greater understanding of the formation of a stable dimer through linker which is an important event for hexamerisation of full length HpDnaB. Also, it reveals surprising details about the linker region of helicase that was earlier reported to be highly flexible and unstructured [37].
Each monomer of NTD consists of seven helices out of which helix1 to helix6 are wrapped to form a globular structure (N-globe) and helix7 extends out from the N-globe. The freely hanging helix7 is highly unstable as it consists of hydrophobic residues on the surface. These hydrophobic residues in the helix7 of two monomers are stabilized by the four helix bundle created all the way through two short peptides which are about 25 residues long helices ( Figure 2D). The electron density suggested that these peptide sequences are similar to helix 7 and MALDI (Matrix-Assisted Laser Desorption/Ionization) analysis of crystals confirmed that these are degraded peptides of helix 7 from NTD itself. The protein degradation was further confirmed by SDS- electrophoresis of crystals ( Figure S1) and MALDI analysis of small degraded peptide found on the 20% SDS-PAGE. Interestingly, the two short peptides are arranged diagonally to each other and interact with the two parallel helices (C-terminal long helix) from two molecules thus forming a four helix bundle ( Figure 2D). This arrangement of four helix bundle allows dimerisation of NTD of HpDnaB. The dimeric interface consists of 29 hydrophobic residues that bury 3410.56 Å 2 of solvent accessible surface area.

Comparison with other structures
The comparison of helicase NTD structures from homologous organisms reveals an unusual arrangement in NTD dimers of HpDnaB. Although, the overall structure of the NTD of HpDnaB is similar to NTDs of MtbDnaB, BstDnaB, TaqDnaB and EcDnaB helicase structure with root-mean-square deviation (RMSD) of 1.42 Å (102 Ca atoms), 1.46 Å (101 Ca atoms), 1.34 Å (93 Ca atoms), 1.45 Å (87 Ca atoms), respectively, even though they share only about 25% sequence identity. An overlay of NTD of HpDnaB on MtbDnaB ( Figure 2E) confirms the propensity of helix7 to be stabilized by forming a helical hairpin. The helix 7 in NTD of HpDnaB is stabilized by the short auto-degraded helices. The overlaying of HpDnaB on NTD of EcDnaB ( Figure 2F) shows that the globular head is compact in both organisms with slight differences in the orientation of secondary structures. As the crystal structure of NTD of EcDnaB lacks the linker region, it does not throw any light on the dimeric organization through linker. Whereas in the solution structure of a little longer construct of NTD of EcDnaB where the linker residues are present, shows a highly disordered linker region [37]. Consequently, the similarity in overall structure is due to the globular domain organization but dimerisation is dependent on the structural assembly and length of linker region that differs in the crystal structures of different NTDs from different organisms.
It has been observed that the NTD of HpDnaB exists as dimers like NTD of MtbDnaB in solution [22], whereas the NTD of EcDnaB exists in monomeric form [37]. However, in crystals, both the NTD of HpDnaB and EcDnaB form dimers, whereas MtbDnaB is observed as a hexamer. Here, E.coli NTD refers to the globular domain only and H. pylori NTD consists of the globular domain along with first helix from linker. However, in MtbDnaB the NTD includes globular domain and two helices from linker region, where these two helices form a-helical hairpin structure. The architectural details of these three structures allow us to understand the differences in dimer formation. The dimers formed by NTD of EcDnaB in crystal structure show interactions through the globular domain alone. While in the solution structure of longer NTD of EcDnaB, the linker does not form helical hairpin rather it is unstructured [37]. In the NTD structure of MtbDnaB, the dimers are formed in a head-to-tail fashion through the four helix bundle, where the helix hairpins from both the molecules are arranged in a head-to-toe manner ( Figure S3). However, the NTD of HpDnaB construct has a half of the helix hairpin (Helix 7) which does not allow the formation of four helix bundle as seen in the structure of MtbDnaB. Stability in NTD of HpDnaB is achieved by forming head to head dimers with four helix bundles at the dimer interface along with short peptides. This clearly shows that the helix hairpin formation in the linker region is highly preferred to enable a biological dimer assembly. The sequence analysis of linker region (helix 7 to 8) reveals, the i(a), i+3(d) and i+4(e) residues in heptad repeats are hydrophobic in nature, which is essential and responsible for four helix bundle formation and this pattern is conserved among different helicases ( Figure S4). Although, the linker region is least conserved among the helicases but it is observed for the first time that the linker sequences (helix 7 and 8) have intrinsic property to form four helix bundles as shown by sequence analysis.

Binding of Primase stimulates helicase activity
To study the HpDnaG-HpDnaB interactions, we had cloned, expressed and purified a putative homolog of DnaG in H. pylori (ORF, Hp0012, Hp strain 26695) for the very first time. The purity and homogeneity of the protein was further checked by SDS-PAGE analysis as shown in ( Figure 3A). The schematic diagram displaying the domain organization in HpDnaG is shown in Figure 4A. The C-terminal domain of HpDnaG binds to the dimeric NTD of DnaB helicase as seen in BstDnaB-DnaG complex structure. Accordingly, we have purified C-terminal primase domain for interaction studies with HpDnaB using surface plasmon resonance (SPR) ( Figure 4B).
The biochemical studies of BstDnaB-DnaG interactions demonstrate that the residues I 119 and I 125 of DnaB establish the foundation for binding of two proteins [38]. These residues fall in the longer helix of helical hairpin of the linker region ( Figure S2) whose counterpart is present in the NTD of HpDnaB used here for crystallization studies. As it is evident from the structure, we were interested to determine whether the unusual dimers of NTD of HpDnaB provides similar platform for binding to CTD of HpDnaG. The sensograms of wtHpDnaB, D2HpDnaB, NTDL and NTD of HpDnaB are shown in Fig. 4B. Sensograms of wtHpDnaB showed a rapid increase in SPR signal during the association phase corresponding to a fast on-rate (K on ), and a gradual reduction in SPR signal during dissociation phase corresponding to a slow off-rate (K off ), suggesting a higher affinity of wtHpDnaB to CTD of HpDnaG (K D = 2.25 mM). On the other hand, a comparatively similar SPR signal accompanied by a faster K on and a faster K off was observed for NTDL of HpDnaB indicating a relatively less affinity towards CTD of HpDnaG (K D = 0.11 mM). A relatively faster K on and K off was observed for NTD of HpDnaB along with a very less SPR signal, indicating that the NTD binds with very less affinity to HpDnaG. The NTD deleted construct (D2HpDnaB) does not binds to CTD of HpDnaG showing the binding takes place through NTD of HpDnaB. Surprisingly, we observed that the NTD of HpDnaB shows very weak binding to HpDnaG primase compared to NTDL and HpDnaB, suggesting that N-terminal domain needs to be in a proper conformation for interaction with HpDnaG primase. These results enhanced our curiosity to determine whether HpDnaG stimulates the HpDnaB helicase activity which is the hallmark of this interaction at the replication fork. For this purpose, the release of radiolabeled unwound oligo from an annealed M13 based helicase substrate was followed in the presence of fixed amount of HpDnaB and increasing amount of HpDnaG as described in Materials and Methods. We find that at low concentration of HpDnaB (20 ng), the release of unwound oligo from the helicase substrate is very weak whereas increasing amount of HpDnaG results in significant stimulation of HpDnaB helicase activity ( Figure 3B). HpDnaG by itself does not show any helicase activity under the same experimental conditions. These results clearly suggest that HpDnaB helicase activity is stimulated by HpDnaG and CTD of HpDnaG (results not shown) in vitro.

Interactions in Helicase-Primase complex
The crystal structure of BstDnaB in complex with CTD of primase [20] was taken as the basic model to understand the helicase-primase interactions. The residues from NTD of DnaB helicase, which participate in primase interactions, come from three different parts of the primary structure ( Figure S2). Two NTDs in the hexameric state of DnaB helicase organize to form a binding surface for primase in three-dimensional space with these three regions. These regions are relatively better conserved in NTD of HpDnaB in contrast with the rest of the N-terminal region ( Figure S2). The critical primase interacting residues from helicase are partially conserved. The differential recognition or binding affinity with primase must be obtained by certain variations in helicase and these variations may be compensated by low homology of the C-terminal domain of primase. For example there are few differences in these three regions, like 64 (L), 65(V/I) and 100(I) which are conserved hydrophobic residues in BstDnaB, EcDnaB, MtbDnaB and TaqDnaB helicases, while these residues are replaced by hydrophilic residues (E, N & E) in HpDnaB. Similarly in primase, three different parts of the sequence come closer in three-dimensional space forming a surface that interacts with NTDs of the two helicases. The primase C-terminal domain has much lower sequence homology compared to the rest of the primase domain or even its interacting counterpart helicase NTD ( Figure S5). Among the four primase sequences, the E.coli DnaG exhibits high diversity and even has an insertion in one of the binding regions. This may explain the weak binding affinity between the E.coli primase and its helicase.

Discussion
The NTD of HpDnaB forms dimers in solution state and in the crystal structure. The NTDs of HpDnaB and EcDnaB exist as dimers in crystal structures while NTD of MtbDnaB is found in the hexameric state. These conformational differences may be primarily due to the relative length of the N-terminal domain construct. The presence of helical hairpin keeps the NTD of MtbDnaB in dimeric state by forming the four helix bundle in head to tail manner, and tertiary interactions of N-terminal domain induce these dimers to assemble as hexamer in crystal state. Helix 7 and 8 belongs to the hairpin forming region of the linker, which is crucial in the positioning of N-terminal lobe during hexamerization of helicase. The degradation of Helix 7 and formation of four helix bundle in our NTD structure and sequence analysis of this region clearly shows that this region has very high tendency to form four helix bundles through hairpin structure to stabilize the hydrophobic residues. The assembly of NTD of HpDnaB through four helix bundle formation and the propensity of helix 7 to form a hairpin structure emphasize on the crucial importance of the linker (specifically helix 7 and 8) in the organization, conformation and function of helicases.
The helical hairpin structure present between the NTD and CTD of helicase plays a critical role in organizing the two NTDs in specific dimeric conformational state by forming four helix bundles. The quaternary interactions of the N-terminal and Cterminal domain stabilize three dimers in the hexameric state ( Figure 5A), which is essential for primase binding. The binding studies show that the CTD of primase binds with a high affinity to the full length helicase and NTDL; poorly with NTD of HpDnaB alone. In the helicase-primase structure of BstDnaB, the two globular heads of NTDs in a hexamer faces one direction to interact with primase whereas in the present structure of NTD of HpDnaB the globular heads are arranged opposite to each other. We presume in NTDL, the helix 7 and 8 from linker region form hairpin which is involved in dimer formation by head to tail arrangement that in turn leads to a hexameric organization, similar to MtbDnaB and that may be responsible for better primase binding. It indicates that the conformational state of NTD of helicase plays a pivotal role in the stability of helicase-primase interactions.
Despite very low sequence homology, the overall globular domain structure of NTD of HpDnaB is similar to NTDs' of EcDnaB, MtbDnaB, BstDnaB and TaqDnaB. Interestingly, the electrostatic potential surface charge distribution on the primase binding surface of helicases show major striking differences in different organisms which may be responsible for the differences in binding stability of DnaB-DnaG complex ( Figure 5B). The primase binding surface of NTD of BstDnaB is exclusively covered by negatively charged side chains at center that must be aiding a strong binding with the CTD of BstDnaG, which is rich in positively charged side chains on the of helicase binding surface. Whereas in E. coli primase binding surface of NTD, the negative charge distribution is very low compared to NTD of BstDnaB, this may be responsible for a transient helicase-primase binding. The primase binding surface of NTD of HpDnaB shows dense negative charge distribution similar to NTD of BstDnaB, which may be responsible for strong HpDnaB-HpDnaG binding. These observations suggest that the helicase and primase proteins display charge dependent interactions.
In conclusion, the structural organization of linker region and binding stability of DnaB-DnaG complex as indicated by surface charge distribution of primase binding surface on helicase shows clearly that HpDnaB differs greatly from EcDnaB despite both belong to gram negative bacteria. Interestingly, the NTD of HpDnaB resembles more with NTDs of MtbDnaB and BstDnaB helicase of gram positive bacteria. The structure of full length DnaB helicase can explain the self-loading phenomenon. Future work of determining the structure of full length HpDnaB/NTDL with primase/primase domain complex will through more light on helicase-primase interactions in H.pylori that may lead to more general understanding of helicase-primase interactions. Figure S1 The MALDI profile of degraded fragment of NTD of HpDnaB. The fragmentation pattern of a peptide m/z = 1895.75 is clearly in agreement with the C-terminal sequence NTIR-EQALEHHHHHH. Inset shows 20% SDS PAGE with lane1, crystal; lane2, Drop without crystal; PM, Protein marker, clearly showing small fragment along with NTD of HpDnaB in crystal. Lane 2, mother liquor shows protein band corresponding to NTD of HpDnaB and some degraded protein. The secondary structures are shown as seen in NTD of HpDnaB. Filled circles represent the residues taking part in helicase-primase interactions as seen in helicase-primase complex structure [20] and filled squares represent the crucial residues essential for maintaining the helicase-primase interactions from biochemical studies in B.stearothermophilus [38].  Figure S4 Sequence alignment of helical hairpin region of DnaB helicase showing the heptad repeat leading to four helix bundle formation. Multiple sequence alignment of helical hairpin region of DnaB from several species (Hp, H. pylori, Bst, B. stearothermohillus; Taq, T. aquaticus; Mtb, M. tuberculosis; Ec, E.coli) are aligned using ClustalW. The (.), (:) and (*) symbols at the bottom of alignment represents the semi conserved, conserved and identical residues in the column. The heptad repeat is labeled as ''abcdefg'' on the top. The i(a), i+3(d) (highlighted with green color) and i+4(e) (highlighted with cyan color) positions are well conserved by hydrophobic residues in helix 8 and C-terminal side of the helix 7 indicating that these induce the four helix bundle formation. In the beginning of helix 7 the i(a) and i+3(d) are conserved with hydrophobic residues indicating that they prefer coiled-coil formation.  [39]. The (.), (:) and (*) symbols at the bottom of alignment represents the semi conserved, conserved and identical residues in the column. Filled circles represent the helicase interacting residues as seen in B.stearothermophilus helicase-primase complex [20]. Found at: doi:10.1371/journal.pone.0007515.s005 (0.93 MB TIF)